Chinese Journal of Chemical Physics  2018, Vol. 31 Issue (4): 457-462

The article information

Li-xue Jiang, Xiao-na Li, Zi-yu Li, Hai-fang Li, Sheng-gui He

H2 Dissociation by Au1-Doped Closed-Shell Titanium Oxide Cluster Anions

Chinese Journal of Chemical Physics, 2018, 31(4): 457-462

http://dx.doi.org/10.1063/1674-0068/31/cjcp1805107

Article history

Accepted on: July 25, 2018
H2 Dissociation by Au1-Doped Closed-Shell Titanium Oxide Cluster Anions
Li-xue Jianga,b,c, Xiao-na Lia,c, Zi-yu Lia,c, Hai-fang Lia,c, Sheng-gui Hea,b,c
Dated: Received on May 15, 2018; Accepted on July 25, 2018
a. State Key Laboratory for Structural Chemistry of Unstable and Stable Species, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China;
b. University of Chinese Academy of Sciences, Beijing 100049, China;
c. Beijing National Laboratory for Molecular Sciences, CAS Research/Education Center of Excellence in Molecular Sciences, Beijing 100190, China
*Author to whom correspondence should be addressed. Xiao-na Li, E-mail:lxn@iccas.ac.cn, Tel.: +86-10-62536990, FAX: +86-10-62559373; Sheng-gui He, E-mail:shengguihe@iccas.ac.cn
Abstract: Dissociation of molecular hydrogen (H2) is extensively studied to understand the mechanism of hydrogenation reactions. In this study, H2 dissociation by Au1-doped closed-shell titanium oxide cluster anions AuTi3O7- and AuTi3O8- has been identified by mass spectrometry and quantum chemistry calculations. The clusters were generated by laser ablation and massselected to react with H2 in an ion trap reactor. In the reaction of AuTi3O8- with H2, the ion pair Au+-O22- rather than Au+-O2- is the active site to promote H2 dissociation. This finding is in contrast with the previous result that the lattice oxygen is usually the reactive oxygen species in H2 dissociation. The higher reactivity of the peroxide species is further supported by frontier molecular orbital analysis. This study provides new insights into gold catalysis involving H2 activation and dissociation.
Key words: Gold    H2 dissociation    Closed-shell anions    Mass spectrometry    Density functional theory calculations
Ⅰ. INTRODUCTION

Oxide supported gold catalysts have been extensively studied owing to their excellent performance in many reactions including hydrogenation [1-4] and formation of hydrogen peroxide [5]. Dissociation of molecular hydrogen (${\rm H}_2$) is usually the rate determining step in catalytic hydrogenation [6, 7] and great efforts have been devoted to investigating the nature of gold catalysis in ${\rm H}_2$ dissociation [8-12]. The perimeter sites [8-13] in between gold particles and oxide supports are generally important for catalytic reactions while the molecular-level mechanism of ${\rm H}_2$ dissociation by the gold-oxide perimeter sites is unclear owing to the complexity of real-life catalysts.

Gas-phase clusters that can be studied under isolated and well controlled conditions are ideal models to uncover the mechanistic details in the related condensed-phase systems [14-20]. Gold-doped heteronuclear metal oxide clusters are being actively studied to understand the molecular-level mechanism of bond activation by the gold-oxide perimeter sites [21-32]. In the reactions of AuCe${\rm O}_2$$^{+} [21], AuNb{\rm O}_4$$^{+}$ [22], ${\rm Au}_2$V${\rm O}_4$$^{+} [23], Au{\rm V}_2$${\rm O}_5$$^{+} [24], and AuV{\rm O}_4$$^{+}$ [25] with ${\rm H}_2$, the separated ion pair ${\rm Au}^+$-O$^{2-}$ can dissociate ${\rm H}_2$ in a heterolytic manner. Some of the studied oxygen-rich clusters including AuNb${\rm O}_4$$^{+} [22], {\rm Au}_2V{\rm O}_4$$^{+}$ [23], and AuV${\rm O}_4$$^{+} [25] contain superoxide ({\rm O}_2$$^{-}$) units. It turns out that the ${\rm Au}^+$-${\rm O}_2$$^{-} ion pair is usually much less reactive than the {\rm Au}^+-O^{2-} ion pair for {\rm H}_2 dissociation. It is noteworthy that the reported metal oxide clusters [33-39] that can dissociate {\rm H}_2 have open-shell electronic structures except for the positively charged cluster AuCe{\rm O}_2$$^{+}$ [21]. Negatively charged cluster anions are usually much less reactive than the cationic counterparts toward reductive molecules [40] and it is rare to identify closed-shell cluster anions that are reactive with stable molecules such as ${\rm H}_2$. Herein, we report that the closed-shell Au${\rm Ti}_3$${\rm O}_7$$^-$ and Au${\rm Ti}_3$${\rm O}_8$$^-$ anions can dissociate ${\rm H}_2$ under thermal collision conditions. Direct H-H bond cleavage by the ${\rm Au}^+$-${\rm O}_2$$^{2-} ion pair has been identified for the reaction of Au{\rm Ti}_3$${\rm O}_8$$^- with {\rm H}_2. This study provides new insights into gold catalysis involving {\rm H}_2 activation and dissociation. Ⅱ. METHODS A. Experimental methods The {\rm Au}_x$$^{48}{\rm Ti}$$_yO_z$$^-$ clusters were generated by laser ablation of a mixed-metal disk compressed with Au and $^{48}$Ti powders (Au:$^{48}$Ti$=$1:1 in molar ratio) in the presence of 0.2% ${\rm O}_2$ seeded in a He carrier gas with a backing pressure of 6.0 standard atmospheres. The clusters of interest (Au${\rm Ti}_3$${\rm O}_7$$^-$ and Au${\rm Ti}_3$${\rm O}_8$$^-$) were mass-selected with a quadrupole mass filter (QMF) and then entered into a linear ion trap (LIT) reactor, where they were confined and cooled by collisions with a pulse of He atoms for around 1.0 ms and then interacted with a pulse of ${\rm H}_2$ or ${\rm D}_2$ for a period of time (1.6-4.7 ms). The temperature of cooling gas (He), reactant gases (${\rm H}_2$, ${\rm D}_2$, or ${\rm N}_2$), and the LIT reactor was around 298 K. The clusters ejected from the LIT were detected by a reflectron time-of-flight mass spectrometer (TOF-MS). The details of running the TOF-MS [41], QMF [39], and LIT [42] can be found in our previous work. The pseudo-first-order rate constants were determined by a least-square fitting procedure [43].

B. Computational methods

Density functional theory (DFT) calculations using the Gaussian 09 [45] program were carried out to study the mechanisms of the reactions between Au${\rm Ti}_3$${\rm O}_{7, 8}$$^-$ and ${\rm H}_2$. The TPSS functional was found to be appropriate to reproduce the dissociation energies of various chemical bonds including Au-H, Au-O, O-H, Ti-O, and O-O [21, 27]. Thus, the TPSS functional was adopted in this work. The TZVP basis set [46] for Ti, H, and O atoms and the D95V basis set [47] combined with the Stuttgart/Dresden relativistic effective core potential (denoted as SDD in Gaussian software) for Au atom were used in the calculations. The relaxed potential energy surface scan was used extensively to obtain good guess structures of intermediates and transition states along the reaction pathways. The transition states were optimized using the Berny algorithm [48]. Intrinsic reaction coordinate calculations [49, 50] were performed to confirm that each transition state connects two appropriate local minima. Vibrational frequency calculations were carried out to check that the intermediates and transition states have zero and only one imaginary frequency, respectively. Furthermore, the single point energies of the low-lying isomers of Au${\rm Ti}_3$${\rm O}_7$$^-$ and Au${\rm Ti}_3$${\rm O}_8$$^-$ clusters were recalculated by the high-level coupled-cluster method with single, double, and perturbative triple excitations (CCSD(T)).

Ⅲ. RESULTS

The TOF mass spectra for the interactions of laser ablation generated and mass-selected Au${\rm Ti}_3$${\rm O}_7$$^-$ and Au${\rm Ti}_3$${\rm O}_8$$^-$ clusters with ${\rm N}_2$, ${\rm H}_2$, and ${\rm D}_2$ are shown in FIG. 1. On the interaction of Au${\rm Ti}_3$${\rm O}_8$$^-$ with 0.03 Pa ${\rm H}_2$ for about 4.7 ms (FIG. 1(b)), signals of ${\rm Ti}_3$${\rm O}_7H^- and {\rm Ti}_3$${\rm O}_8$H$^-$ as well as Au${\rm Ti}_3$${\rm O}_6$$^-$ and Au${\rm Ti}_3$${\rm O}_7$$^-$ could be identified. When high pressure ${\rm H}_2$ was used (0.60 Pa, FIG. 1(c)), the signal intensity of Au${\rm Ti}_3$${\rm O}_6$$^-$ increased significantly while that of Au${\rm Ti}_3$${\rm O}_7$$^-$ nearly disappeared, indicating that Au${\rm Ti}_3$${\rm O}_6$$^-$ was due to the secondary reaction of the resulting Au${\rm Ti}_3$${\rm O}_7$$^-$ with ${\rm H}_2$ (reaction (1a) and then reaction (2a)). The generation of ${\rm Ti}_3$${\rm O}_7H^- and {\rm Ti}_3$${\rm O}_8$H$^-$ ions in Au${\rm Ti}_3$${\rm O}_8$$^-$+${\rm H}_2$ suggests formation of neutral species AuOH (reaction (1b)) and AuH (reaction (1c)), respectively.

 ${\rm Au}{\rm Ti}_3{\rm O}_8^-+{\rm H}_2 \rightarrow {\rm Au}{\rm Ti}_3{\rm O}_7^-+{\rm H}_2{\rm O}$ (1a)
 ${\rm Au}{\rm Ti}_3{\rm O}_8^-+{\rm H}_2 \rightarrow {\rm Ti}_3{\rm O}_7{\rm H}^-+{\rm AuOH}$ (1b)
 ${\rm Au}{\rm Ti}_3{\rm O}_8^-+{\rm H}_2 \rightarrow {\rm Ti}_3{\rm O}_8{\rm H}^-+{\rm AuH}$ (1c)
 ${\rm Au}{\rm Ti}_3{\rm O}_7^-+{\rm H}_2 \rightarrow {\rm Au}{\rm Ti}_3{\rm O}_6^-+{\rm H}_2{\rm O}$ (2a)
 FIG. 1 Time-of-flight mass spectra for the reactions of mass-selected Au${\rm Ti}_3$${\rm O}_8$$^-$ and Au${\rm Ti}_3$${\rm O}_7$$^-$ with ${\rm N}_2$ ((a) and (e)), ${\rm H}_2$ ((b), (c), and (f)), and ${\rm D}_2$ ((d) and (g)) are shown. ${\rm Au}_x$${\rm Ti}_y$${\rm O}_z$$^- and {\rm Au}_x$${\rm Ti}_y$${\rm O}_zX (H or D)^- species are labeled as x, y, z and x, y, zX respectively. The time periods for reactions Au{\rm Ti}_3$${\rm O}_8$$^-$$+$${\rm H}_2 and Au{\rm Ti}_3$${\rm O}_7$$^-$$+$${\rm H}_2 are about 4.7 and 1.7 ms, respectively. Products {\rm Ti}_3$${\rm O}_7$H$^-$, ${\rm Ti}_3$${\rm O}_8H^-, Au{\rm Ti}_3$${\rm O}_6$$^-, and Au{\rm Ti}_3$${\rm O}_7$$^- did not appear in the reaction of Au{\rm Ti}_3$${\rm O}_8$$^- with {\rm N}_2 (FIG. 1(a)), indicating that these signals are due to the chemical reaction rather than collision-induced dissociation. The assignment of the product ions was confirmed by the isotopic-labeling experiment (FIG. 1(d)). On the interaction of mass-selected Au{\rm Ti}_3$${\rm O}_7$$^- with 0.01 Pa {\rm H}_2 (FIG. 1(f)), the product of Au{\rm Ti}_3$${\rm O}_6$$^- was identified. This provides further support that the Au{\rm Ti}_3$${\rm O}_6$$^- in FIG. 1(b) was due to the secondary reaction of the resulting Au{\rm Ti}_3$${\rm O}_7$$^- with {\rm H}_2. With respect to Au{\rm Ti}_3$${\rm O}_8$$^-+{\rm H}_2 (FIG. 1(c)), much lower {\rm H}_2 pressure (0.01 Pa) and shorter reaction time (~1.7 ms) were used in Au{\rm Ti}_3$${\rm O}_7$$^-+{\rm H}_2 to have significant signal depletion of the reactant cluster (Au{\rm Ti}_3$${\rm O}_7$$^-). This means that Au{\rm Ti}_3$${\rm O}_7$$^-+{\rm H}_2 is much faster than Au{\rm Ti}_3$${\rm O}_8$$^-+{\rm H}_2, which well rationalizes the disappearance of Au{\rm Ti}_3$${\rm O}_7$$^- in FIG. 1(c). In addition to Au{\rm Ti}_3$${\rm O}_6$$^-, {\rm Ti}_3$${\rm O}_7$H$^-$ was also observed in the reaction of Au${\rm Ti}_3$${\rm O}_7$$^-$ with ${\rm H}_2$ (reaction (2b)). The production of ${\rm Ti}_3$${\rm O}_7H^- was confirmed by the isotopic-labeling experiment (FIG. 1(g)).  {{\rm AuT}}{{{\rm i}}_{{\rm 3}}}{{\rm O}}_7^ - {{\rm + }}{{{\rm H}}_{{\rm 2}}} \to {{\rm T}}{{{\rm i}}_{{\rm 3}}}{{{\rm O}}_{{\rm 7}}}{{{\rm H}}^{{\rm - }}}{{\rm + AuH}} (2b) The pseudo-first-order rate constants (k_1) for the reactions of Au{\rm Ti}_3$${\rm O}_{7, 8}$$^- with {\rm H}_2 and {\rm D}_2 can be well fitted (FIG. S1 and FIG. S2 in supplementary materials) and the results are presented in Table Ⅰ. It turns out that about 57% of experimentally generated Au{\rm Ti}_3$${\rm O}_8$$^- could react with {\rm H}_2 and the reactive component for Au{\rm Ti}_3$${\rm O}_7$$^- is about 78%. It has been often reported that a cluster can have different isomeric structures with very different reactivity [26, 51]. For the reactive component of Au{\rm Ti}_3$${\rm O}_8$$^-, the k_1 value is (0.061\pm0.018)\times$$10^{-10}$ c${\rm m}^3$$\cdot$${\rm molecule}^{-1}$$\cdot$${\rm s}^{-1}$ with the reaction efficiency [52] of 0.41%. The kinetic isotope effect $k_1$(Au${\rm Ti}_3$${\rm O}_8$$^-$+${\rm H}_2$)/$k_1$(Au${\rm Ti}_3$${\rm O}_8$$^-$+${\rm D}_2$) is 1.2. The $k_1$ value for Au${\rm Ti}_3$${\rm O}_7$$^-$+${\rm H}_2$ is (5.8$\pm$1.7)$\times$$10^{-10} c{\rm m}^3$$\cdot$${\rm molecule}^{-1}$$\cdot$${\rm s}^{-1} with the reaction efficiency of 39%. The kinetic isotope effect k_1(Au{\rm Ti}_3$${\rm O}_7$$^-+{\rm H}_2)/k_1(Au{\rm Ti}_3$${\rm O}_7$$^-+{\rm D}_2) is 1.2. Table Ⅰ Branching ratios (BRs), total rate constants (k_1, in unit of 10^{-10} c{\rm m}^3$$\cdot$${\rm molecule}^{-1}$$\cdot$${\rm s}^{-1}), reaction efficiency (\Phi), and kinetic isotope effect (KIE) for the reactions of Au{\rm Ti}_3$${\rm O}_8$$^- and Au{\rm Ti}_3$${\rm O}_7$$^- cluster anions with {\rm H}_2 or {\rm D}_2. The DFT calculated reaction pathways for {\rm H}_2 dissociation by Au{\rm Ti}_3$${\rm O}_8$$^- and Au{\rm Ti}_3$${\rm O}_7$$^- are shown in FIG. 2 and FIG. 3, respectively. The cluster isomers of Au{\rm Ti}_3$${\rm O}_8$$^- and Au{\rm Ti}_3$${\rm O}_7$$^- have been carefully calculated in our previous study [27] and both of the clusters have closed-shell electronic structures. In the lowest-lying isomer of Au{\rm Ti}_3$${\rm O}_8$$^- (IS1, FIG. 2(a)), the gold atom is bridgingly bounded with one lattice oxygen (O^{2-}) and one peroxide unit ({\rm O}_2$$^{2-}$ with O-O bond length of 148 pm). Such gold atom can hardly trap ${\rm H}_2$ (the binding energy is smaller than 0.04 eV). In contrast, the isomers with the terminally bonded gold atom (IS2 and IS3) can trap and react with ${\rm H}_2$. The positively charged gold (natural charge: +0.39 e) in Au${\rm Ti}_3$${\rm O}_8$$^-$ (IS2) can anchor ${\rm H}_2$ tightly with a binding energy of 0.92 eV (I1, FIG. 2(b)). The H-H bond in I1 is significantly activated because the H-H bond length increases from 74 pm in free ${\rm H}_2$ to 89 pm in I1. In the next step, the gold atom delivers the attached ${\rm H}_2$ to the ${\rm O}_2$$^{2-} unit and then H-H bond cleaves with a barrier of 0.13 eV (I1\rightarrowTS1\rightarrowI2, FIG. 2(b)). The formed OOH species is an important intermediate in the condensed phase [11] and the O-O bond cleavage occurs with a barrier of 0.55 eV (I2\rightarrowTS2\rightarrowI3). The subsequent steps (from I3 to I7) mainly involve the H atom transfer with a nearly downhill pathway to form a very stable intermediate I7 (-3.46 eV) with two O-H units. The I7 can transform into I8 that can evaporate the AuOH unit favorably to produce the experimentally observed major product ion {\rm Ti}_3$${\rm O}_7$H$^-$.

 FIG. 2 DFT calculated potential energy profiles of (a) transformation among the low-lying isomers of Au${\rm Ti}_3$${\rm O}_8$$^-$ and (b) and (c) the reaction Au${\rm Ti}_3$${\rm O}_8$$^-$$+$${\rm H}_2$ on the singlet state. The relative energies for IS1-IS4, intermediates (I1-I11), transition states (TS1-TS10), and products are in unit of eV. Structures of Au${\rm Ti}_3$${\rm O}_8$$^-$ and I1-I11 are shown. Bond lengths are given in unit of pm. P1: ${\rm Ti}_3$${\rm O}_7H^-$$+$AuOH, P2: Au${\rm Ti}_3$${\rm O}_7$$^-$$+$${\rm H}_2$O, P3: ${\rm Ti}_3$${\rm O}_8H^-$$+$AuH. The values in the square brackets (a) are the single-point energies calculated at the CCSD(T) level.
 FIG. 3 DFT calculated potential energy profiles of (a) transformation between two low-lying isomers of Au${\rm Ti}_3$${\rm O}_7$$^-$ and (b) the reaction Au${\rm Ti}_3$${\rm O}_7$$^-$$+$${\rm H}_2$ on the singlet spin state. The relative energies for cluster isomers IS5, IS6, intermediates (I12-I15), transition states (TS11-TS13), and products are in unit of eV. Bond lengths are given in unit of pm. P4: ${\rm Ti}_3$${\rm O}_7H^-$$+$AuH, P5: Au${\rm Ti}_3$${\rm O}_6$$^-$$+$${\rm H}_2$O. The values in the square brackets (a) are the single-point energies calculated at the CCSD(T) level.

Starting from intermediate I2, the reaction complex can also form I9 (I2$\rightarrow$TS8$\rightarrow$I9) that can evaporate a water molecule to generate the experimentally observed weak product ion Au${\rm Ti}_3$${\rm O}_7$$^-$ (FIG. 1(b)). The relative energy of the critical transition state (TS3: -0.94 eV) in the pathway to generate ${\rm Ti}_3$${\rm O}_7H^- is significantly lower than the energy of TS8 (-0.65 eV) and Au{\rm Ti}_3$${\rm O}_7$$^- can further react with {\rm H}_2 to form {\rm Ti}_3$${\rm O}_7$H$^-$, so the ${\rm Ti}_3$${\rm O}_7H^- signal can be much stronger than the Au{\rm Ti}_3$${\rm O}_7$$^- signal in FIG. 1(b). Starting from intermediate I1, the reaction complex can also form I10 (I1\rightarrowTS9\rightarrowI10, FIG. 2(c)) that can cleave the H-H bond through the {\rm Au}^+-O^{2-} ion pair (I10\rightarrowTS10\rightarrowI11). The intermediate I11 can evaporate AuH to form the experimentally observed product ion {\rm Ti}_3$${\rm O}_8$H$^-$. From Rice-Ramsperger-Kassel-Marcus theory, the conversion rate of I10$\rightarrow$TS10$\rightarrow$I11 should be faster than that of I2$\rightarrow$TS8$\rightarrow$I9, which is consistent with the result that the ${\rm Ti}_3$${\rm O}_8H^- signal is stronger than the Au{\rm Ti}_3$${\rm O}_7$$^- signal in FIG. 1(b). The lowest-lying isomer of Au{\rm Ti}_3$${\rm O}_7$$^- (IS5, FIG. 3(a)) at the CCSD(T) level corresponds to the reactive isomer of Au{\rm Ti}_3$${\rm O}_7$$^-. FIG. 3(b) shows that the generation of {\rm Ti}_3$${\rm O}_7$H$^-$+AuH (P4, -0.95 eV) is slightly less facile than the generation of Au${\rm Ti}_3$${\rm O}_6$$^-$+${\rm H}_2$O (P5, -0.97 eV). However, generation of P5 involves tight transition states TS13 and TS14 should be entropically less favorable than generation of P4, which is consistent with the experimental result (FIG. 1(f)). It should be pointed out that the experiments (FIG. 1) indicated that the reactive component of Au${\rm Ti}_3$${\rm O}_7$$^-$ is much more reactive than that of Au${\rm Ti}_3$${\rm O}_8$$^-$ in the reaction with ${\rm H}_2$. However, the current DFT results (FIG. 2 and FIG. 3) cannot well explain this difference. It is possible that the IS1 or IS4 (FIG. 2) is the reactive component of Au${\rm Ti}_3$${\rm O}_8$$^-$ and the ${\rm H}_2$ adsorption and Au-O bond breaking that can also lead to formation of the intermediate I1 may be the bottleneck of Au${\rm Ti}_3$${\rm O}_8$$^-$+${\rm H}_2$ (FIG. S6 in supplementary materials).

Ⅳ. DISCUSSION

In the field of gas-phase studies, different oxygen species have been proposed to activate and dissociate ${\rm H}_2$. In the reactions of Os${\rm O}_4$$^{+} [38] and {\rm V}_4$${\rm O}_{10}$$^{+} [39] with {\rm H}_2, the atomic oxygen radical anion {\rm O}^- is the reactive oxygen species (ROS). Recent studies on the reactions of gold-doped heteronuclear metal oxide clusters highlighted the importance of lattice oxygen {\rm O}^{2-} in direct H-H bond dissociation [21-25], paralleling the behavior of {\rm H}_2 dissociation in the condensed phase systems [54, 55]. Furthermore, indirect participation of the superoxide radical ({\rm O}_2$$^{-}$) has also been proposed [22, 23, 25]. Herein, in the reaction of Au${\rm Ti}_3$${\rm O}_8$$^-$ with ${\rm H}_2$, direct H-H bond cleavage by the ${\rm Au}^+$-${\rm O}_2$$^{2-} ion pair with peroxide unit has been identified and this pathway is more facile than that by the {\rm Au}^+-O^{2-} ion pair. To further understand the mechanism of {\rm H}_2 dissociation by Au{\rm Ti}_3$${\rm O}_8$$^-, the natural charges on the Au atom, the two hydrogen atoms, and the {\rm O}_2$$^{2-}$ unit during ${\rm H}_2$ dissociation (I1$\rightarrow$TS1$\rightarrow$I2) are monitored and provided in FIG. 4(a). In the intermediate I2, the positively charged H atom in the OOH unit (H2, +0.50 e) and the negatively charged H atom (H1, -0.26 e) in the AuH unit are good indicators of heterolytic cleavage of the H-H bond, which is the most common mechanism of ${\rm H}_2$ dissociation by gold-containing species in both of the gas-phase and the condensed-phase systems [2, 11, 21, 54, 55]. Note that the charge on the ${\rm O}_2$$^{2-} unit is nearly constant during {\rm H}_2 dissociation (-0.66 e in I1 vs. -0.72 e in I2), which is consistent with the same O-O bond length (149 pm) in I1 and I2, pointing to the fact that a proton ({\rm H}^+) is attached on the {\rm O}_2$$^{2-}$ unit in I2.

 FIG. 4 DFT calculated (a) natural charge on the Au atom, two H atoms, and the ${\rm O}_2$$^{2-} unit during {\rm H}_2 dissociation (I1\rightarrowTS1\rightarrowI2, see FIG. 2(b)) and (b) molecular orbital for Au{\rm Ti}_3$${\rm O}_8$$^- (left) and intermediate I1 (right). The structure of I2 is provided in (a). H and L in (b) represent the highest occupied and the lowest unoccupied molecular orbital, respectively. The orbital energies are in unit of eV. The frontier molecular orbital analysis (FIG. 4(b)) shows that for Au{\rm Ti}_3$${\rm O}_8$$^-, the lowest unoccupied molecular orbital (LUMO) is on the AuO moiety which can be the electron-acceptor. In contrast, the highest occupied molecular orbital (HOMO) is on the {\rm O}_2$$^{2-}$ unit which can be the electron-donor. Upon ${\rm H}_2$ adsorption, the LUMO of Au${\rm Ti}_3$${\rm O}_8$$^-$ accepts the $\sigma$ electrons of ${\rm H}_2$ and the ${\rm H}_2$ $\sigma$$^* orbital becomes the LUMO of the intermediate I1. The HOMO of I1 is almost identical to the HOMO of Au{\rm Ti}_3$${\rm O}_8$$^-. Furthermore, the HOMO-LUMO gap increases significantly upon {\rm H}_2 adsorption, from 0.09 eV in Au{\rm Ti}_3$${\rm O}_8$$^- to 2.15 eV in Au{\rm Ti}_3$${\rm O}_8$${\rm H}_2$$^-$ (I1), indicating that the ${\rm H}_2$ adsorption can greatly stabilize the cluster, as reported previously [58]. In I1, the peroxide ${\rm O}_2$$^{2-} and the lattice oxygen O^{2-} contribute to the HOMO and HOMO-1, respectively, which is consistent with the DFT result (FIG. 2) that the {\rm Au}^+-{\rm O}_2$$^{2-}$ ion pair is more reactive than the ${\rm Au}^+$-O$^{2-}$ ion pair. This gas-phase study parallels well the condensed-phase result that the adsorbed molecular ${\rm O}_2$ at the perimeter sites between gold nanoparticles and Ti${\rm O}_2$ supports can reduce the barrier for ${\rm H}_2$ dissociation [11].

Ⅴ. CONCLUSION

In conclusion, the Au${\rm Ti}_3$${\rm O}_7$$^-$ and Au${\rm Ti}_3$${\rm O}_8$$^-$ clusters can react with and dissociate ${\rm H}_2$ under thermal collision conditions. To the best of our knowledge, this is the first report that the cluster anions with closed-shell electronic structures can activate and dissociate the very stable di-hydrogen. The H-H bond is cleaved in a heterolytic manner through cooperation of the Lewis acid-base ion pair of ${\rm Au}^+$-O$^{2-}$ or ${\rm Au}^+$-${\rm O}_2$$^{2-}. In the reaction of Au{\rm Ti}_3$${\rm O}_8$$^- with {\rm H}_2, the ion pair involved with the peroxide unit ({\rm Au}^+-{\rm O}_2$$^{2-}$) is more reactive than that with the lattice oxygen (${\rm Au}^+$-O$^{2-}$) for ${\rm H}_2$ dissociation. This is in contrast with the common consideration that the lattice oxygen is the reactive oxygen species to dissociate ${\rm H}_2$ in the reaction systems of metal oxides. New insights are provided into gold catalysis involving ${\rm H}_2$ activation and dissociation.

Supplementary materials: Detailed description of experimental and theoretical methods and additional theoretical results are given.

Ⅵ. ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (No.21573246, No.21773253, and No.21773254), the Beijing Natural Science Foundation (2172059). Xiao-na Li thanks the Youth Innovation Promotion Association, Chinese Academy of Sciences (2016030).

Reference
 [1] M. Boronat, and A. Corma, Langmuir 26 , 16607 (2010). DOI:10.1021/la101752a [2] Z. P. Liu, C. M. Wang, and K. N. Fan, Angew. Chem. Int. Ed. 45 , 6865 (2006). [3] X. Zhang, H. Shi, and B. Q. Xu, Angew. Chem. Int. Ed. 44 , 7132 (2005). DOI:10.1002/(ISSN)1521-3773 [4] C. Mohr, H. Hofmeister, J. Radnik, and P. Claus, J. Am. Chem. Soc. 125 , 1905 (2003). DOI:10.1021/ja027321q [5] P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely, and G. J. Hutchings, Chem. Commun. , 2058 (2002). [6] G. C. Bond, and P. B. Wells, Adv. Catal. 15 , 91 (1964). [7] F. Meemken, and A. Baiker, Chem. Rev. 117 , 11522 (2017). DOI:10.1021/acs.chemrev.7b00272 [8] I. X. Green, W. Tang, M. Neurock, and J. T. Yates Jr., Acc. Chem. Res. 47 , 805 (2014). DOI:10.1021/ar400196f [9] K. Sun, M. Kohyama, S. Tanaka, and S. Takeda, J. Phys. Chem. C 118 , 1611 (2014). DOI:10.1021/jp4099254 [10] D. Widmann, E. Hocking, and R. J. Behm, J. Catal. 317 , 272 (2014). DOI:10.1016/j.jcat.2014.06.011 [11] I. X. Green, W. Tang, M. Neurock, and J. T. Yates Jr., Angew. Chem. Int. Ed. 50 , 10186 (2011). DOI:10.1002/anie.201101612 [12] T. Fujitani, I. Nakamura, T. Akita, M. Okumura, and M. Haruta, Angew. Chem. Int. Ed. 48 , 9515 (2009). DOI:10.1002/anie.200905380 [13] X. Y. Liu, A. Q. Wang, T. Zhang, and C. Y. Mou, Nano Today 8 , 403 (2013). DOI:10.1016/j.nantod.2013.07.005 [14] K. R. Asmis, Phys. Chem. Chem. Phys. 14 , 9270 (2012). DOI:10.1039/c2cp40762k [15] S. M. Lang, and T. M. Bernhardt, Phys. Chem. Chem. Phys. 14 , 9255 (2012). DOI:10.1039/c2cp40660h [16] P. J. Roach, W. H. Woodward, A. W. Castleman Jr, A. C. Reber, and S. N. Khanna, Science 323 , 492 (2009). DOI:10.1126/science.1165884 [17] C. Chi, H. Qu, L. Meng, F. Kong, M. Luo, and M. Zhou, Angew. Chem. Int. Ed. 56 , 14096 (2017). DOI:10.1002/anie.201707898 [18] R. Burgert, H. Schnockel, A. Grubisic, X. Li, S. T. Stokes, K. H. Bowen, G. F. Gantefor, B. Kiran, and P. Jena, Science 319 , 438 (2008). DOI:10.1126/science.1148643 [19] H. Schwarz, S. Shaik, and J. Li, J. Am. Chem. Soc. 139 , 17201 (2017). DOI:10.1021/jacs.7b10139 [20] M. Krstic, Q. Jin, G. N. Khairallah, R. A. J. O'Hair, and V. Bonacic-Koutecky, Chemcatchem 10 , 1173 (2018). DOI:10.1002/cctc.201701594 [21] J. H. Meng, and S. G. He, J. Phys. Chem. Lett. 5 , 3890 (2014). DOI:10.1021/jz502057n [22] L. X. Jiang, X. N. Li, H. F. Li, Z. X. Zhou, and S. G. He, Chem. Asian J. 11 , 2730 (2016). DOI:10.1002/asia.201600144 [23] X. H. Zhou, Z. Y. Li, T. M. Ma, and S. G. He, J. Phys. Chem. C 120 , 10452 (2016). DOI:10.1021/acs.jpcc.6b02515 [24] Y. Zhang, Z. Y. Li, Y. X. Zhao, H. F. Li, X. L. Ding, H. Y. Zhang, and S. G. He, J. Phys. Chem. A 121 , 4069 (2017). DOI:10.1021/acs.jpca.7b02435 [25] X. P. Zou, Z. Y. Li, X. N. Li, L. N. Wang, H. F. Li, T. M. Ma, and S. G. He, Top. Catal. 61 , 28 (2018). DOI:10.1007/s11244-017-0860-7 [26] X. N. Li, Z. Y. Li, H. F. Li, and S. G. He, Chem. Eur. J. 22 , 9024 (2016). DOI:10.1002/chem.201600451 [27] Y. X. Zhao, X. N. Li, Z. Yuan, Q. Y. Liu, Q. Shi, and S. G. He, Chem. Sci. 7 , 4730 (2016). DOI:10.1039/C6SC00539J [28] Z. Y. Li, H. F. Li, Y. X. Zhao, and S. G. He, J. Am. Chem. Soc. 138 , 9437 (2016). DOI:10.1021/jacs.6b03940 [29] L. N. Wang, Z. Y. Li, Q. Y. Liu, J. H. Meng, S. G. He, and T. M. Ma, Angew. Chem. Int. Ed. 54 , 11720 (2015). DOI:10.1002/anie.201505476 [30] X. N. Li, Z. Yuan, and S. G. He, J. Am. Chem. Soc. 136 , 3617 (2014). DOI:10.1021/ja412608b [31] H. Himeno, K. Miyajima, T. Yasuike, and F. Mafun, J. Phys. Chem. A 115 , 11479 (2011). DOI:10.1021/jp202125g [32] Z. Yuan, X. N. Li, and S. G. He, J. Phys. Chem. Lett. 5 , 1585 (2014). DOI:10.1021/jz500509j [33] S. G. Ard, J. J. Melko, O. Martinez Jr, V. G. Ushakov, A. Li, R. S. Johnson, N. S. Shuman, H. Guo, J. Troe, and A. A. Viggiano, J. Phys. Chem. A 118 , 6789 (2014). DOI:10.1021/jp5055815 [34] M. Zhou, C. Wang, J. Zhuang, Y. Zhao, and X. Zheng, J. Phys. Chem. A 115 , 39 (2011). DOI:10.1021/jp109498b [35] M. Zhou, C. Wang, Z. H. Li, J. Zhuang, Y. Zhao, X. Zheng, and K. Fan, Angew. Chem. Int. Ed. 49 , 7757 (2010). DOI:10.1002/anie.v49:42 [36] M. F. Ryan, A. Fiedler, D. Schröder, and H. Schwarz, J. Am. Chem. Soc. 117 , 2033 (1995). DOI:10.1021/ja00112a017 [37] D. E. Clemmer, N. Aristov, and P. B. Armentrout, J. Phys. Chem. 97 , 544 (1993). DOI:10.1021/j100105a005 [38] K. K. Irikura, and J. L. Beauchamp, J. Am. Chem. Soc. 111 , 75 (1989). DOI:10.1021/ja00183a014 [39] Z. Yuan, Y. X. Zhao, X. N. Li, and S. G. He, Int. J. Mass Spectrom. 354-355 , 105 (2013). DOI:10.1016/j.ijms.2013.06.004 [40] Y. X. Zhao, Q. Y. Liu, M. Q. Zhang, and S. G. He, Dalton Trans. 45 , 11471 (2016). DOI:10.1039/C6DT01246A [41] X. N. Wu, B. Xu, J. H. Meng, and S. G. He, Int. J. Mass Spectrom. 310 , 57 (2012). DOI:10.1016/j.ijms.2011.11.011 [42] Z. Yuan, Z. Y. Li, Z. X. Zhou, Q. Y. Liu, Y. X. Zhao, and S. G. He, J. Phys. Chem. C 118 , 14967 (2014). DOI:10.1021/jp5040344 [43] X. P. Zou, L. N. Wang, X. N. Li, Q. Y. Liu, Y. X. Zhao, T. M. Ma, and S. G. He, Angew. Chem. Int. Ed. 57 , 10989 (2018). DOI:10.1002/anie.v57.34 [44] G. Kummerlöwe, and M. K. Beyer, Int. J. Mass Spectrom. 244 , 84 (2005). DOI:10.1016/j.ijms.2005.03.012 [45] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. M. Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox Gaussian 09, Revision A. 1, Wallingford, CT: Gaussian, Inc., (2009). [46] A. Schafer, C. Huber, and R. Ahlrichs, J. Chem. Phys. 100 , 5829 (1994). DOI:10.1063/1.467146 [47] D. Andrae, U. Häußermann, M. Dolg, H. Stoll, and H. Preu, Theor. Chim. Acta 77 , 123 (1990). DOI:10.1007/BF01114537 [48] H. B. Schlegel, J. Comput. Chem. 3 , 214 (1982). DOI:10.1002/(ISSN)1096-987X [49] C. Gonzalez, and H. B. Schlegel, J. Chem. Phys. 90 , 2154 (1989). DOI:10.1063/1.456010 [50] C. Gonzalez, and H. B. Schlegel, J. Phys. Chem. 94 , 5523 (1990). DOI:10.1021/j100377a021 [51] R. Kretschmer, M. Schlangen, and H. Schwarz, Angew. Chem. Int. Ed. 52 , 6097 (2013). DOI:10.1002/anie.201300900 [52] G. Gioumousis, and D. P. Stevenson, J. Chem. Phys. 29 , 294 (1958). DOI:10.1063/1.1744477 [53] J. I. F. Steinfeld, J. S. Hase, W. L., 2nd ed., New Jersey: Prentice-Hall, (1999). [54] G. Lu, P. Zhang, D. Q. Sun, L. Wang, K. B. Zhou, Z. X. Wang, and G. C. Guo, Chem. Sci. 5 , 1082 (2014). DOI:10.1039/C3SC52851K [55] B. Yang, X. M. Cao, X. Q. Gong, and P. Hu, Phys. Chem. Chem. Phys. 14 , 3741 (2012). DOI:10.1039/c2cp23755e [56] A. Corma, M. Boronat, S. Gonzalez, and F. Illas, Chem. Commun. , 3371 (2007). [57] B. Yoon, H. Häkkinen, U. Landman, A. S. Worz, J. M. Antonietti, S. Abbet, K. Judai, and U. Heiz, Science 307 , 403 (2005). DOI:10.1126/science.1104168 [58] N. K. Jena, K. R. S. Chandrakumar, and S. K. Ghosh, J. Phys. Chem. C 113 , 17885 (2009). DOI:10.1021/jp906295n

a. 中国科学院化学研究所分子动态与稳态结构国家重点实验室, 北京 100190;
b. 中国科学院大学, 北京 100049;
c. 中国科学院分子科学科教融合卓越创新中心, 北京分子科学国家实验室, 北京 100190